The gypsy moth, Lymantria dispar (L.), was introduced in the United States around 1869 around Boston. Since that time, it has spread slowly to the North, West, and South, and its current range includes all of the Northeastern States, as well as portions of Michigan, Ohio, West Virginia, Virginia, and North Carolina (Liebhold et al. 1992). Over the next 50 years it is likely that the gypsy moth will expand its range to include much of the Southeast and Midwest. Over the last 100 years, defoliation by this insect has caused considerable damage to several of the most valuable tree species in the Northeast, particularly the oaks (Quercus spp). To prepare for this pest in currently uninfested regions, there is a need to identify the tree species that the gypsy moth is likely to attack.

Several studies have attempted to classify the susceptibility of various tree species to defoliation by the gypsy moth. Perhaps the most meaningful studies have been those that quantified defoliation levels on individual tree stands with high-density populations of gypsy moth (Campbell and Sloan 1977; Gansner and Herrick 1985). While these studies provide useful information on the susceptibility of different tree species to the gypsy moth, the geographic scope of such empirical field data is limited to the Northeastern United States. Information about the susceptibility of tree species that are not found in this region is available only from laboratory studies (e.g., Forbush and Fernald 1896; Mosher 1915; Miller and Hanson 1989a).

In reviewing gypsy moth feeding trials, Montgomery (1991) pointed out that it is difficult to compare these studies because virtually no two used the same techniques. In some trials, gypsy moths were confined to the foliage of two or more species; in others, there were no choices. Further, some studies based evaluations of host suitability solely on growth (i.e., weight gain), others on survival, and still others on the propensity of gypsy moths to feed on the foliage. In fact, The suitability of a tree species in the field is a combination of both choice and performance (Montgomery 1991).

Apart from the difficulties in comparing data from non-standardized trials, there is a need to combine information on tree suitability into a single source. The objective of this report was to tabulate the results of field observations and laboratory feeding trials to obtain a generalization of the suitability of tree species. For each species we developed a ranking (1-3) of foliage suitability. For species with extensive tests in the literature, this rank represents a summary of these tests. For tree species for which there are few or no reports, the rank is a "best guess" based on the results of field and laboratory tests of closely related species.